> Hi, > > I would just like to know. When downloading a file from a site say for > example a link like so: http://host1.example.com/test.zip and the file > has a size say 300kb (and timestamp 2006/01/01 ?) and then downloading > the same file from another (mirror) site say > http://host2.example.com/test.zip with the filesize also at 300kb and > the same timestamp Will squid redownload the file? Or will it be > "clever" enough to say that I've already downloaded that file, I'll send > you the cached version? Do timestamps matter? Does squid just check > the filesizes and filenames? > There's no way for SQUID to know, it matches the same object, SQUID only uses the complete url to decided on that. (MD5 - checksum) Whether cached copies will be returned, if already accessed, (from the same 'uniq' server); also depends on freshness info provided by the webserver ,about the returned object. What if, in both cases , the webserver issues a 'Cache-control: no-cache ' header. Then you question collapes by definition. M.
